Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Context Triage

    1

    Generate a local repository manifest and Git status report to prevent agent hallucinations during coding tasks.

    Free

    Security scanned
    Context Triage

    Context Triage

    Example session with this skill installed

    Initialize context for the current repository before I start refactoring the auth module.

    • Read your context and instructions
    • Compiled the context triage

    I have scanned the repository and generated a context manifest. The AGENT_CONTEXT.md file is now available as a reference for the current session.

    Repository Context Manifest

    AttributeValue
    StatusSuccess
    Manifest Path./AGENT_CONTEXT.md
    Files Found24
    Current Branchfeature/auth-refresh
    Pending Changessrc/auth/service.py, tests/test_auth.py
    Untracked Filesconfig/secrets.yaml.tmp

    Detected Structure Summary

    • src/auth/: 6 files (Modified)
    • src/middleware/: 4 files (Clean)
    • scripts/: 2 files (Clean)
    • tests/: 12 files (Modified)

    Next steps

    1. Review AGENT_CONTEXT.md to confirm all relevant auth files were detected.
    2. Proceed with the refactoring, focusing on the src/auth/ directory.
    3. Commit the current pending changes before applying new logic to avoid conflicts.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Stop your AI agent from hallucinating file paths. Context Triage forces the agent to scan the local directory and check Git status before it starts writing code. It generates a clean AGENT_CONTEXT.md file that gives the agent a perfect map of the repository.

    Why use it?

    Prevents the agent from editing files that are already modified.

    Gives a clean inventory of all relevant Python/JS/TS files.

    Zero setup, no API keys required.

    Makes your paid skills (like Repo Map or Sprint Ready) work better.

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for LLM Ops.

    Does not read file content or perform deep code analysis. Requires Git to be initialized in the target directory for status checks.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before buying a skill.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Free to use

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 2 days ago

    • Free with an Agensi account

    Software
    Git

    The manifest is generated locally and printed to stdout — nothing is uploaded. Run it before your agent starts a coding task; the output tells the agent what's in the repo and what's changed, so it stops guessing at file paths.

    Permissions

    Terminal / Shell
    Read Files
    Write Files
    Listed2 days ago
    Updatedtoday

    What's inside

    Frequently Asked Questions

    This skill prevents 'hallucinations' where agents invent non-existent file paths or attempt to modify files that have uncommitted local changes, ensuring your agent always works from the actual current state of your repository.

    Context Triage is a Python-based utility compatible with any environment that allows local script execution, including popular AI IDEs like Cursor and Windsurf, as well as custom-built agentic workflows.

    The skill requires a local installation of Python and Git; it works by scanning your directory structure and metadata without needing to send your actual code to an external server for the manifest generation.

    You receive the core Python utility script, integration instructions for major AI coding assistants, and a template for the AGENT_CONTEXT.md ground-truth file.

    The skill focuses on file mapping and Git status only; it does not read the internal content of your files, keeping the context window focused on the repository structure rather than overwhelming it with raw code.

    Yes, the manifest can be scoped to the entire repository or a specific sub-path, making it ideal for navigating large monorepos where you only need the agent to focus on a specific module.

    New in LLM Ops

    Free